Heads I Win; Tails Don't Count:
The actual value of abstinence
by Stephen Hanson
The following article is from Free Inquiry magazine, Volume
One frequently hears the claim that "Abstinence is the only
100-percent-certain way to avoid pregnancy and sexually transmitted diseases
(STDs)." Proponents of abstinence-only sex education say it, the Bush
administration says it, and even advocates of other forms of birth control say
it. In many ratings of the effectiveness of birth-control methods, abstinence is
listed as being as effective as hysterectomy and more effective than vasectomy:
100 percent. What is remarkable about all this is that this statement is not
true in almost any sense. As I will show, the only sense in which it is true is
one that provides no particular motivation for someone to choose abstinence over
other forms of contraception and disease prevention. Understood in any way other
than this trivial one, it is either extremely misleading or false.
First, the sentence is not strictly true: even perfect abstinence does not
provide one with 100 percent certainty of avoiding STDs, because most are
transmittable in other fashions (e.g., accidental needle sticks, etc.). It is
also possible to become pregnant while abstinent under some definitions of the
term, because semen can occasionally contact the vagina even without vaginal
penetration. But perhaps this misses the desired point: one can be certain of
not getting pregnant or getting an STD through sexual contact as long as one
defines abstinence as avoiding any action that allows any exchange of bodily
fluids, including contact between semen and female genitalia. Then, one can
maintain that abstinence provides 100-percent certainty of avoiding pregnancy or
sexual transmission of STDs.
The claim is true in this fashion: if one does not perform a given action,
then there can be no consequences of performing that action. So we can hold the
above claim to mean, "For each case where one could perform a sexual act, not
performing one will avoid the consequences of performing sexual acts," which is,
of course, true.
But that is not the way that the claim is used. Abstinence is put forth not
as a single action, but rather as a method of birth control. To judge the
effectiveness of abstinence in this respect, we need to judge not only whether
the individual act of abstaining from sexual intercourse provides the stated
100-percent benefits, but also whether choosing to use abstinence as a method of
birth control provides such results. The difference between the two is
significant, as it points to the fundamental difference between the way
abstinence and all other birth-control methods are treated when determining
their success rates.
In the context of assessing the effectiveness of a birth-control method, one
must differentiate between a method's theoretical effectiveness and its
"use-effectiveness" or "typical use-effectiveness." The theoretical
effectiveness of a method of birth control is a "perfect world" calculation-no
one ever forgets to take a pill, condoms never slip and are always put on before
any genital contact, etc. It is therefore of little interest to the average
person except as an intellectual exercise. What we want to know, when choosing a
method of birth control, is how well it will work in the real world to prevent
pregnancy. It is here that abstinence is treated radically differently than any
other method of birth control.
The first reason for this is practical and yet still points to the problem in
making this comparison. The use-effectiveness of a given form of contraception
depends, of course, at least in part, on the rate of sexual intercourse, though
this is not always made clear. The chances of becoming pregnant while utilizing
a given method of birth control are greater the more frequently one has sexual
intercourse. One source notes that "Researchers presume that married women have
coitus three times weekly and their statistics assume [this] 'married model.'"1
Thus, unless one averages thrice-weekly intercourse, the actual
use-effectiveness of one's birth control will differ from that calculated on the
"married model." If one has sex more frequently, the success rate will be lower;
if less frequently, the success rate will be higher.
Whether this model is accurate for most marriages, or whether it is used for
other lists, it is clearly not used for any abstinence model. For in determining
the effectiveness of abstinence, but only of abstinence, no sexual intercourse
is presumed. Immediately the comparison is suspect, if only for the reason that
we are comparing apples and oranges-the success rate of the pill assumes sex
thrice a week, while the success rate of abstinence presumes none. Naturally,
both the theoretical and use-effectiveness of various other methods of birth
control, including none whatsoever, increase dramatically when one assumes no
Since the definition of abstinence is "no sexual intercourse," it is perhaps
unfair to criticize someone reporting the success rate of abstinence on grounds
that they have not presumed regular sexual intercourse. But this points to the
real problem, which is that the "100-percent-success rate" is only theoretical.
If we assume one always succeeds in avoiding sex, then we can fairly assume the
success rate of abstinence to be 100 percent (with the exceptions noted above).2
But the use-effectiveness-the success rate in the real world-of persons trying
to use abstinence as their method of birth control is far less than 100 percent
because they do not always succeed in avoiding sexual intercourse.
This is not news, I suspect. People trying to be abstinent can and do fail,
and sometimes pregnancy results. What is astonishing is how little attention is
paid to this fact. On list after list of effectiveness of birth-control methods,
abstinence is ranked at the top, or given special mention because of its
100-percent-success rate. This pride of place is confusing and quite possibly
dangerous. Because of the difference in what is being measured, to trumpet the
value of abstinence relative to other methods is at best deceptive, at worst
irrational. Since nothing of what I have said above is new or surprising, and
since the use-effectiveness of abstinence is surely significantly less than its
theoretical effectiveness, how does abstinence continue to be labeled as the
most effective method of birth control on the grounds of its theoretical
effectiveness, even when the use-effectiveness of other methods is reported?
One reason may be that the use-effectiveness rate of abstinence, or exactly
how we should calculate it, is unclear. Many sources give no answer at all; and
though some recent research has attempted to properly determine abstinence's
use-effectiveness, its results have not made their way into the popular
understanding.3 While we know a great deal about use-effectiveness
rates for other methods of contraception, a recent review of the literature on
sex education argued, "We still know very little about abstinence and how often
adolescents [or, presumably, any other persons] fail to use abstinence
consistently."4 Given this, how can anyone label abstinence as
100-percent effective, even in theory, without a very serious caveat to the
effect that we have no clear idea how effective it is in practice?
I suggest that the reason is a "Heads I win, tails don't count" way of
thinking. When a couple attempts to avoid having sex as a means of birth control
and succeeds, they have properly used abstinence as their method of birth
control, and lo and behold, they achieve 100-percent success at preventing
pregnancy. If, on the other hand, they attempt to use abstinence as their method
of birth control but end up having sexual intercourse, they have not abstained
and thus haven't used abstinence. They are then included in the group of people
who used no method of birth control. The success rate of abstinence remains 100
This kind of language game sustains the tenet that "abstinence is the only
sure way," but it is intellectually dishonest. The sentence "Abstinence is the
only 100-percent-certain way to avoid pregnancy and STDs" is true only in a
trivial way that also applies to all other forms of birth control-assuming no
sexual activity, there will be no impact from sexual activity, regardless of
one's chosen method of protection. And when one tries to find out the number of
women who become pregnant while trying to use abstinence as birth control, one
encounters at best the misleading language about failing to be abstinent-and
therefore not using abstinence. Whether one takes them regularly or not, if one
is depending upon a birth-control pill to prevent pregnancy, one is using the
pill as one's method of birth control-this is the difference between theoretical
and use-effectiveness. The same difference applies to abstinence as a
birth-control method, and it's time we discover and properly report those
numbers as well. If we are going to talk about slipping condoms, failures to
take pills on time, etc. as relevant to how effective these methods of birth
control are-and we should-then we had better talk as well about failures of will
in avoiding one of the most fundamental drives that humans have.
February 24, 2003.
2. Actually, the success rate is not 100 percent, but rather undefined. The
success rate can be read as a ratio of "times one did not get pregnant from
sexual intercourse" over "times one had sexual intercourse." If the denominator
is zero, the ratio is undefined, even if the numerator is zero as well. Hanno
Bulhof pointed this out to me.
3. Steven D. Pinkerton, "A Relative Risk-Based, Disease-Specific Definition
of Sexual Abstinence Failure Rates." Health Education and Behavior 28,
no. 1 (February 2001).
4. Clara S. Haignere, Rachel Gold, and Heather J. McDaniel, "Adolescent
Abstinence and Condom Use: Are We Sure We Are Really Teaching What Is Safe?"
Health Education and Behavior 26, no. 1 (February 1999): 43-54.
Stephen Hanson is in the Department of Social Sciences at McNeese State
University in Louisiana.